So I have been thinking about trying to put together a chronological list of yoyo statistics. I am well aware of the scope of the project and will actively be looking for people to assist in the project if there is enough support for the idea. I am thinking about having a list of 4 to 5 categories per year and going back as far as records can be found (well aware large time gaps will happen the further back we get).
Currently I am thinking about having one poll category per year as well and that category I think could be something to the tune of like most influential yoyo of that year?
I also would like to have top yoyo in sales from boutique retailers, so not sure if this is some data I could work with the folks at yye to get a start on?
Would like to get some topic suggestions as well to fill the last slots! Not sure if this is a realistic undertaking but I keep thinking how neat it would be to have a list like this. Let me know what you guys think, and let me know if you want to take part in the project if I can get the idea to take off!
So you are basically looking to digitise the mind of @jhb8426
Haha in a way. I think this more is a combination of my cabin fever and my love for data statistic and fun facts all finding a way to coexist
I love this idea
There would be royalties involved.
Like do you want to be knighted? Lol this is definitely an idea that will take a lot of peoples input so we shall see if there is enough interest to push forward with it.
@jhb8426 is already royalty, knighthood would be a step down. Cold hard cash works though.
Cash is king for sure!
Here’s something I’ve been thinking about for quite a while that would go along with this effort nicely. If I could get my hands on a large set of pictures of yoyos, each correctly labeled by brand & model it would be pretty easy for me to train a classifier which when given a picture of a yoyo would give you its best guesses of what it is. Ideally I’d have pictures of them from all kinds of angles and in different settings.
I do a lot of work in machine learning and I have the hardware resources to train the model. If I have a clean dataset this won’t be hard to do. We’d need to find somewhere to host it so everybody could make use of it.
I would love this
In case you’re doing this I would propose to all mine mentions of yoyos and building a data driven recommender system for yoyos. Maybe even mapping language like floaty to actual weight distributions and specs